GPT-6 Sol now leads DeepSecBench's Pareto frontier:
• Highest benchmark score
• 80% lower run cost than the runner-up (GPT-6 Astra)
vercel.com/ai-gateway/leader…
Vercel Sandbox now has memory observability.
𝚟𝚌 𝚖𝚎𝚝𝚛𝚒𝚌𝚜 𝚟𝚎𝚛𝚌𝚎𝚕.𝚜𝚊𝚗𝚍𝚋𝚘𝚡.𝚖𝚎𝚖𝚘𝚛𝚢_𝚞𝚜𝚎𝚍_𝚋𝚢𝚝𝚎𝚜
You and your agent can see P75, P95, and average usage for a single project or your entire team. vercel.com/changelog/vercel-…
Introducing Vercel's public Bug Bounty Program.
It combines our private and OSS bounties into a single @Hacker0x01 program, open to anyone. vercel.com/blog/the-vercel-b…
Unlimited Vercel Blob stores on every plan.
Starting today there's no limit on how many stores you can have. A store per customer, project, or job now works at any scale.
vercel.com/changelog/unlimit…
Gemini 3.8 TTS is live on AI Gateway.
100+ languages. Try both models:
• 𝚐𝚘𝚘𝚐𝚕𝚎/𝚐𝚎𝚖𝚒𝚗𝚒-𝟹.𝟾-𝚏𝚕𝚊𝚜𝚑-𝚕𝚒𝚝𝚎-𝚝𝚝𝚜
• 𝚐𝚘𝚘𝚐𝚕𝚎/𝚐𝚎𝚖𝚒𝚗𝚒-𝟹.𝟾-𝚏𝚕𝚊𝚜𝚑-𝚝𝚝𝚜
vercel.com/changelog/gemini-…
Vercel Sandbox now has persistent storage with Drives, in public beta on every plan.
▪︎ Store agent workspaces, data, models, deps
▪︎ Read snapshots across parallel sandboxes
▪︎ Mount up to four Drives per sandbox
▪︎ Up to 16 TiB per Drive
vercel.com/changelog/drives-…
GPT-6 Sol and Luna are live on AI Gateway.
Half the price of GPT-5.6, while building on Astra's advancements.
• 𝚘𝚙𝚎𝚗𝚊𝚒/𝚐𝚙𝚝-𝟼-𝚜𝚘𝚕
• 𝚘𝚙𝚎𝚗𝚊𝚒/𝚐𝚙𝚝-𝟼-𝚕𝚞𝚗𝚊
vercel.com/changelog/gpt-6-s…
Opus 5.5 is now on AI Gateway, with support for fast mode.
Stronger agentic coding and knowledge work, at lower cost and faster speed than previous Opus versions.
𝚊𝚗𝚝𝚑𝚛𝚘𝚙𝚒𝚌/𝚌𝚕𝚊𝚞𝚍𝚎-𝚘𝚙𝚞𝚜-𝟻.𝟻
vercel.com/changelog/claude-…
MiMo V2.6 models are live on AI Gateway.
Trained for coding, agents, vision & cyber.
• 𝚡𝚒𝚊𝚘𝚖𝚒/𝚖𝚒𝚖𝚘-𝚟𝟸.𝟼-𝚙𝚛𝚘
• 𝚡𝚒𝚊𝚘𝚖𝚒/𝚖𝚒𝚖𝚘-𝚟𝟸.𝟼-𝚏𝚕𝚊𝚜𝚑
• 𝚡𝚒𝚊𝚘𝚖𝚒/𝚖𝚒𝚖𝚘-𝚟𝟸.𝟼-𝚙𝚛𝚘-𝚞𝚕𝚝𝚛𝚊𝚜𝚙𝚎𝚎𝚍
vercel.com/changelog/mimo-v2…
Two new ways to call Jev through AI Gateway, alongside the AI SDK:
▪︎ TypeSafe-compatible API, same eval calls
▪︎ HTTP API for any language or framework
Read more ↓
vercel.com/changelog/ai-gate…
Microsoft Teams is now on Vercel Connect.
Your apps and agents run as a Teams bot. People mention it in channels or message it directly, and your code replies as the bot.
vc connect create microsoft-teams
Use your team's private packages or design system in v0 with one shared npm credential.
Set 𝙽𝙿𝙼_𝚃𝙾𝙺𝙴𝙽 or 𝙽𝙿𝙼_𝚁𝙲 once as a sensitive shared environment variable on Vercel. The credential is injected at the network layer. The model and agent never see it.
vercel.com/changelog/v0-now-…
Jev was adopted faster than any other model in AI Gateway history.
In the first day, @typesafeai reached ~13% of teams, 2x the GPT-5.6 family and 6x Fable 5.1.
We collaborated with @HacktronAI to reproduce the libheif vulnerability in Next.js and responsibly disclose it to the maintainer.
Learn about the dependency chain and fix ↓
vercel.com/blog/reproducing-…
Agents browsing your site can use WebMCP tools instead of clicking through your UI.
𝚖𝚌𝚙-𝚑𝚊𝚗𝚍𝚕𝚎𝚛 𝟸.𝟸.𝟶 adds experimental WebMCP support. Allowlist tools and add one script tag. In-browser agents call them as the signed-in user, with no OAuth flow.
vercel.com/changelog/webmcp-…
GPT-Live 1 is on AI Gateway.
• Full duplex audio model: listens while speaking, so you can interrupt.
• Client delegation: run any text model in the background.
𝚘𝚙𝚎𝚗𝚊𝚒/𝚐𝚙𝚝-𝚕𝚒𝚟𝚎-𝟷
vercel.com/changelog/gpt-liv…
GLM 5.3 FlashX is live on AI Gateway.
Same vision, tool use, and 1M context as GLM 5.3 Flash, at higher TPS.
𝚣𝚊𝚒/𝚐𝚕𝚖-𝟻.𝟹-𝚏𝚕𝚊𝚜𝚑𝚡
vercel.com/changelog/glm-5-3…
You and your coding agent can now use the CLI to deploy static artifacts in less than one second.
▲ ~/ vercel deploy ./my-static-site --prod
vercel.com/changelog/sub-sec…